Game Design Slight rework to Agility

In my eyes, the current current agility dodge system is flawed and leads to many people having misgivings about high agility low defence builds. As of right now I'm using a build with around 140 Agility (~75% dodge chance) and 0 defence, resulting in the following ehp calculations:

• Effective HP: 31923
• Effective HP (no agi): 10565

If i were to instead use a build with 140 defence (~68% damage reduction) 0 agility i would get the following ehp calculations:

• Effective HP: 32782
• Effective HP (no agi): 32782

While both of these cases have similar ehp values, it is safe to assume most people would rather use the latter build.

Why?
In this case, due to the randomness associated with agility dodges, a 75% chance to dodge instead feels like a % chance to get hit twice in a row (or more) with only 10k ehp.

In an example where you take 4 hits (same damage), the odds that 2 or more of them don't get dodged in a row is around 13.3% (this would increase for longer hit "chains", which is usually what happens in game during, for example, a lootrun challenge or a raid challenge room). Unfortunately, taking quick successive hits at 10k ehp is usually a death sentence to most archetypes.

Meanwhile, with a defence heavy approach those 4 hits would always result in taking equally damaging hits for a total equivalent damage of 1.28 hits at 10k ehp. This results in a much better survival chance since each hit will be relatively weaker and the burst damage window will be much greater, allowing for more time to react appropriately by either escaping from danger or healing

My proposal
The only other game similar to Wynncraft that i have any experience with would be Path of Exile (an ARPG) which has similar defensive mechanics with amour (defence, outside the scope of this post) and evasion (agility).

The evasion system works in a very intuitive/interesting manner, it still works based off of a % chance of dodging a hit but instead utilizes an entropy based calculation system for whether a hit gets dodged or not, resulting in a much more deterministic approach to dodging.

You can read more about it on the game's wiki but the basics of the system would work as following:

• When first attacked (or if the last attack was a while ago) a random entropy value is rolled between 0-99.
• To calculate whether an attack hits or not, add (100 - dodge chance) to your entropy value.
• If the resulting entropy value exceeds 100, subtract 100 from the value and the attack registers as a hit, otherwise it is dodged.

Here's an example using 70% dodge chance:

The player is attacked, their starting entropy counter is randomly assigned between 0-99, in this case 83 (fairly unlucky).​

• The first hit adds 30 to the counter, raising it to 100 or greater, and therefore doesn't get dodged. 100 is subtracted and the entropy is now at 13.
• The second hit adds another 30 to the counter, for a total of 43 (dodged).
• Third hit adds 30, entropy of 73 (dodged).
• Fourth hit adds 30, entropy of 103, doesn't get dodged and entropy is now at 3.
• The 5th, 6th, and 7th hit would be a dodge
• The 8th hit doesn't get dodged

The player runs away for some time (say for example like 3+ seconds) without getting hit, so a new starting entropy counter will be rolled on the next attack. (ensuring the first attack in a chain is always random and not a guaranteed dodge from being hit by the previous attack that may have happened long ago).​

What results from this is that having 75% dodge chance, which basically means 1 in 4 chance to get hit, will result in you always getting hit exactly once out of every 4 consecutive attacks.

Pros and Cons
- Only way I see this being implemented is if max dodge chance is reduced to bellow 75%, in which case we would be gaining less ehp from agility but that ehp would certainly feel more "real" than the current version (while also making sure it doesn't outclass Defence in terms of a purely defensive option)

+ Heavily diminishes the feeling of randomness associated with high agility playstyles, making survivability more dependent on player input, and less so on a slot machine gambling system, by increasing the delay between each undodged hit, giving the player more time to react defensively.

+ Brings Agility closely in line with Defence in terms of consistency, not so much in performance.

+ Somewhat maintains the High-Risk High-Reward playstyle of heavy agility investment playstyles, while also making it more deterministic (keep in mind you cannot really control whether the dodge is going to trigger (or not) on a heavier hit or a weaker one).

+ Less people driven away by the perceived luck/gambling heavy playstyle of high Agility builds.

+Effectively similar to the current system when applied to low agility investment (<40% dodge chance) since the dodges will still be mostly random and not as deterministic, ensuring this only really affects high agility investment builds

+ Not overly complicated to implement (I think?), just call the random function once and then add/compare/subtract while storing time since last hit (negligible server impact i would assume).

- dev time ;-;

 

I think the biggest factor that your suggestion overlooks is that the dmg reductions are not held in a vacuum. One of the biggest advantages that agi as an element has over def is walk speed. Walk speed is not only amazing QoL, but it's also MASSIVELY impactful in many boss fight encounters. While not strictly necessary/helpful in all fights, the indirect effect that walk speed has on a build's longevity/ability to take on many bosses/enemies should not be ignored, which i think this post has done. Otherwise I do like the idea of less rng and more consistency with dodges, so im not totally opposed to this suggestion!

 

Yeah i understand that adding consistency to agility (high ws, very strong items) could very well make it a little too strong, which (if it does turn out to be a problem) could just as easily be "balanced" by making max dodge like 70% or slightly increasing the damage from dodged hits

 

Also in terms of Raid Bosses I do not believe this change to be overly impactful.

In NotG you can either:

• Stand in a corner and face tank the boss, in which case the added benefit of high ws becomes kind of null and this change merely bring agility more in-line with defence when it comes to surviving any burst damage you might receive.
• Play the boss as normal, in which case you probably won't be hit too often in quick succession (remember that the first hit is always completely random).

In NoL you take so much chip damage from random sources that this change practically affects nothing.

In TCC most attacks are either high damage large AoE with a very large time window between each (meaning your dodges are once more random just as it is currently implemented) or something more bursty like the balls/laser which hit enough times and fast enough to guarantee you will get hit atleast twice (very quick succession) if you do not try to dodge... which is already fairly trivial with high ws.

In TNA I could definitively see this being a bit too strong with dodge values at or above 75% (guarantees 3 dodged hits following an undodged hit) but this should otherwise be mitigated by the fact that many of the attacks deal damage in separate instances (you can, for example, dodge only half or a third of the damage in an attack) or consist of short bursts of many damaging hits (would functionally work the same as the currently implemented agility dodge), and with high enough ws most of the times there will be some window of time between each hit that would reset the entropy counter and make the next hit random.

The main target of this change wasn't so much the bosses and more so the challenge room where the constant hits from mobs, which in time might induce death from not dodging 3+ times in a short span of time, these would instead be less random since this system works by limiting how many times you unluckily get hit in a row and how many times you can luckily dodge in a row (no more dodging 10 times in a row). By lowering the entropy counter reset time to something like 1 or 2 seconds you could probably also make this change less impactful in boss fights (less deterministic guaranteed dodges).

 

Edited the pros and cons section to emphasise that some changes would have to be made so agility doesn't ever become as good or better than defence in terms of sheer survivability.

 

I disagree because like you said in your previous post, it mostly behaves the same anyway because the first hit is still random and you can’t really predict dodges on successive hits anyway. All it does is prevent really unlucky or lucky scenarios, which everyone agrees is bullshit anyway. I really don’t see how this change would make it better than defense due to the randomness it still has. The EHP of agility is still the same as before on average, so if the devs think having 75% dodge chance is fine, then I see no reason to nerf it then.

This is just how Agility should have always been and given the Discord poll a lot of people feel the same too (I haven’t seen a post from a non-staff get over 100 likes in that section before).

Also add a poll. (And this should be in the General Suggestions instead bc more people will see it than Feedback).

 

Reason that Agility feels so underwhelming as a defensive option right now is that, even if it says on paper that you are getting 30k+ ehp, you will eventually be hit 2-3 times in a short succession while at 10k ehp and die (it has alot of variance in performance), if we reduced said variance there is no way it should still keep that 30k+ ehp, Defence should be the strongest choice when investing in survivability and it's items clearly showcase that, implementing this change with no nerfs would put Agility far closer to Defence in terms of consistency while also keeping the high ehp gain from agility (keep in mind as of now one point in Agility gives like 95%+ of the ehp of one point in Defence so the disparity you mention isn't really significant). This would make Agility far too competitive with Defence in terms of survivability while still keeping all the advantages inherent to air items (high mobility, high damage), completely ruining the game's balance.

Despite this proposed change still having some rng, you will *never* get hit twice or more in a row, signifiantly improving your survivability despite not increasing the ehp.

By reducing the variance in Agility (making it nearly as consistent as Defence) we have to compensate by nerfing the overall survivability somehow, which is most easily done by lowering the ehp.

You basically go from "High ehp but it feels inconsistent and 'fake'..." to "Medium ehp but it feels more consistent and 'real'...".

PS: added a poll

 

But that’s assuming you always get hit within 1 or 2 seconds, which isn’t true at all. And you still can’t really choose which hits you want to proc either, so relying entirely on agility for defense can still easily kill you unlike an equivalent amount of defense. So even with this change, I would still say defense is overall better than agility in terms of taking hits. Besides, overall you’re still taking the same theoretical amount of damage averaged out in a fight as before. Therefore, agility having other indirect perks like walkspeed is fine bc it’s EHP is relatively unaffected.

It’s also important to consider that indirect bonuses of fire items too if we are also considering Air’s. Fire items also give direct HP bonuses and health regen buffs, so even if Agility and Defense were the same in terms of EHP by themselves, defense would still be better in terms of EHP due to its indirect item buffs.

Plus it seems very hyperbolic to say the game’s balance would be completely ruined if the agility cap stays at 76.8% instead of 75%… At that point you’re better off putting some remainder skill points into defense than you are trying to reaching 150 in agility.

 

I refuse to use any build that relies entirely on agility, purely because one or two shot kills are common enough in the endgame that even if it was a 90% chance, sooner or later the odds say you're gonna get obliterated. This would dramatically reduce the odds of that happening, as it would only be true oneshots you'd have to be as worried about.